evolutionary morphology, platyrrhine evolution, and...

THE ANATOMICAL RECORD 294:1955–1974 (2011)

Evolutionary Morphology, PlatyrrhineEvolution, and Systematics

ALFRED L. ROSENBERGER1,2,3*1Department of Anthropology and Archaeology, Brooklyn College,

The City University of New York, Brooklyn, New York2New York Consortium in Primatology (NYCEP), The Graduate Center,

The City University of New York, New York, New York3The American Museum of Natural History, Department of Mammalogy,

New York, New York

ABSTRACTThis special volume of the Anatomical Record focuses on the

evolutionary morphology of New World monkeys. The studies range fromthree-dimensional surface geometry of teeth to enamel ultrastructure; fromcranioskeletal adaptations for eating leaves and seeds to the histology oftaste bud proxies; from the architecture of its bones to the mechanoreceptorsof the tail’s skin; from the physical properties of wild foods to the feeding bio-mechanics of jaws and skull; from the shapes of claws and fingertips, and ofelbows, to the diversity and morphology of positional behavior; from the vom-eronasal organ and its biological roles to links between brains, guts, sociality,and feeding; from the gum-eating adaptations of the smallest platyrrhines tothe methods used to infer how big the largest fossil platyrrhines were. Theydemonstrate the power of combining functional morphology, behavior, andphylogenetic thinking as an approach toward reconstructing the evolution-ary history of platyrrhine primates. While contributing new findings pertain-ing to all the major clades and ecological guilds, these articles reinforce theview that platyrrhines are a coherent ecophylogenetic array that differenti-ated along niche dimensions definable principally by body size, positionalbehavior, and feeding strategies. In underlining the value of character analy-sis and derived morphological and behavioral patterns as tools for decipher-ing phylogenetic and adaptational history, doubts are raised about acompeting small-bore morphological method, parsimony-based cladistic stud-ies. Intentionally designed not to enlist the rich reservoir of platyrrhine evo-lutionary morphology, an empirical assessment of the costs incurred by thisresearch stratagem reveals inconsistent, nonrepeatable, and often conflictingresults. Anat Rec, 294:1955–1974, 2011. VVC 2011 Wiley Periodicals, Inc.

Keywords: platyrrhines; New World monkeys; functionalmorphology; phylogeny; adaptive radiation;systematics

Knowing more and more about less and less maymean that relationships are lost and that the grandpattern and great processes of life are overlooked.

George Gaylord Simpson (1944:xxvii)

Tempo and Model in Evolution

The impetus for this special volume of the AnatomicalRecord grew out of an excited conversation with one ofmy good coworkers who has spent his entire professionallife studying primate—especially New World monkey—

morphology, enlightened by personal experience dissect-ing primates, collecting their fossils, chasing monkeys in

*Correspondence to: Dr. Alfred L. Rosenberger, Department ofAnthropology and Archaeology, Brooklyn College, The City Univer-sity of New York. New York. E-mail: [email protected]

Received 15 September 2011; Accepted 16 September 2011

DOI 10.1002/ar.21511Published online 1 November 2011 in Wiley Online Library(wileyonlinelibrary.com).

VVC 2011 WILEY PERIODICALS, INC.

the forest, and performing experiments in the lab. Myrhetorical question was: ‘‘Is morphology dead?’’ His an-swer, paraphrasing, was: ‘‘It’s never been more alive.Look at all the students going to the museum, codingcharacters.’’ To which I replied: ‘‘It’s not about making aGrey’s Anatomy of primate morphology. What more havewe learned about the animals since all that began?’’

The articles assembled here, which focus on platyr-rhines, one of the major primate radiations, will hope-fully demonstrate that primate morphology is anythingbut dead, nor is it in abeyance in the era of genomics.Primate morphology is more vibrant than ever, as itshould be. We have refined the questions, deployedimproved methods, are mastering advanced technologiesthat probe more deeply into the body and its workings,and continue to discover new fossils that revealunexpected morphologies which expand the anatomicalboundaries circumscribed by extant forms. However, asSimpson warned, it is the relationships among phenom-ena, the hardest thing to understand and the mostimportant, that remains at risk when investigative agen-das are very particularized and synthesis does not followbasic research. At that nodal point, where observationpivots toward meaning, the conversations we havebehind the printed word indicates one whole version ofmorphological inquiry evinces controversy anduncertainty.

I refer to the morphology behind systematics, morespecifically to the anatomical basis and methods of phy-logeny reconstruction. One needs neither footnotes norreferences to validate what is obvious in the pages of theleading journals that publish on primate evolution. Putbluntly: morphological articles dealing with phyloge-netics are overwhelmingly biased toward studies thatseem to use traits as if they are the parts of inanimatesystems, without any reference to a feature’s organic lifeand history. It is an ironic intellectual twist. The obso-lete ‘‘descriptive anatomy’’ on which old-school system-atics was built, and which deserved a much needed shotof fresh blood and thinking, has morphed into a nonde-scriptive anatomy beholden to ‘‘character states’’ that, byintention, are normally construed to be blind to hypothe-ses, suppositions, and assumptions. The result is adefleshed, binary anatomy. In adopting it, this newschool of systematics atomizes and transmographiesstructural complexity in an effort to satisfy the twingoals of minimizing subjectivity and maximizing samplesize, that is, the number of traits used and the numberof taxa consulted, en route to producing machine-basedcladograms. Is this a good thing? Some say, Yes; otherssay, No.

Evolutionary morphology, the theme of this volume, isneither a new construct nor a new label (e.g., Szalay,2000). I use the term here simply as an expressionmeant to avoid the artificiality of separating the func-tional properties of characters from their phylogeneticproperties. Both aspects are intrinsic to morphologyalthough studying either facet often requires differenttools and approaches. However, my contention is thatthere is more explanatory power in a synthetic morphol-ogy where form is brought together with function andphylogeny conceptually. I try to demonstrate this whileintroducing and contextualizing the articles delivered inthe following pages, pointing out where an evolutionaryor functional morphological perspective has been impor-

tant in deciphering and explaining key points in platyr-rhine phylogeny and history, as elucidated by anatomy.This approach is a break with the old schools of taxon-omy, when nonfunctional paradigms dominated platyr-rhine systematics. It is also the antipode of algorithmic,new-school systematics, an explicitly nonfunctionalistapproach to phylogeny reconstruction. In my view, noother group of living primates illustrates better the ben-efits of a synthetic evolutionary morphology approach tosystematics than the platyrrhines.

Following Szalay (e.g., 2000) and others, one can alsomake a distinction between evolutionary morphologyand functional morphology as it is now practiced in thatthe former is inherently concerned with transformation.Functional morphology is surely steeped in evolution, asthe explanations we have for ‘‘how things work’’ can berelated to selective pressure and differential survivalunder ecological circumstances. Thus, folivory is seen asadaptive when one lives in a world where leaves arepotential food and theoretically beneficial to survival.However, functional morphology does not address howadaptations like folivory arose. This represents a distinctset of questions and an approach that differs from thebiomechanical. Tracing the origins of a trait or complexrequires phylogenetic thinking, but it would be littlemore than a sterile exercise if it was not infused withpossible explanations of ‘‘why’’ it happened. Evolutionarymorphology, as the articles presented herein often show,gets us closer to establishing that crucial relationshipbetween How and Why, which Simpson (1944) wouldsurely have seen as part of the grand pattern and greatprocess behind platyrrhine evolution.

Concerning the articulation of morphology and sys-tematics, to be clear at the outset: I do not wholly rejectcomputer-driven analyses of morphology designed toprobe cladistics on philosophical or theoretical grounds.Phylogeny reconstruction is a difficult scientific chal-lenge and all helpful approaches are welcome. However,now that we are about 20 years into the process ofapplying them to platyrrhines, there is ample empiricalevidence to allow an objective assessment of these stud-ies, most of which have used the parsimony method (i.e.,PAUP; Swofford, 2002). This leads me to question howwell they have been implemented as well as theircapacity to perform as intended (see Matthews andRosenberger, 2008; Rosenberger, 2010b). I also wonderhow to evaluate the many inconsistencies they turn out(see below), and if they actually live up to the criticaltheoretical requirement that they produce explicitly(pure) cladistic results. Do the results of these studiesjustify the underlying proposition that we should assumeas little as possible about the organization and evolutionof anatomical characters so as not to contaminate theexercise by making supposedly unnecessary a prioripostulates?

Such a viewpoint seems merely to shift the menu ofcomplex scientific assumptions—all of science involvesassumptions—from the input stage to the analytical.Because of it, hard earned information, hypotheses andcausal explanations concerning functional morphology,adaptation and evolutionary history tend to be rejectedas valuable evidence. They rarely inform the input rou-tine, character selection and coding, and seem to holdlittle sway when they challenge output in the form of analternative cladistic hypothesis. The atomization of

1956 ROSENBERGER

morphology required by the algorithmic approach stripsmorphology of its effective and historical patterns, lead-ing to an objectification which may have an effect oppo-site of what is intended—fabricating more noise. For inany orderly structural system, the information contentof each unit is proportional to the size of the unit withinthe system and the level of its integration with otherparts. Is there any doubt a primate’s 10 toenails will sayless about history than a whole primate foot functionallyinterpreted? Or, that the laptop’s many keys say lessabout the manufacturer where it originated than theone motherboard to which each is connected? Ultimately,I am doubtful that some of the logical bases and expecta-tions of the algorithmic agenda has been proven true,that more small-bore observations are commensuratewith higher quality research, more accurate results, andthe virtue of real objectivity.

In posing this contrast between approaches to phyloge-netics, the platyrrhines serve as an object lesson in thevalue of integrating cladistics with an evolutionary, func-tional morphology approach to systematics. These themesspeak to some of the relationships Simpson (1944) fearedwere falling between the cracks. More specifically, I takethe view that the platyrrhines illustrate how characteranalysis can contribute to a holistic assessment of evolu-tionary history that accounts for phylogeny and adapta-tion within the same theoretical and methodologicalframework. One of the chief aims of this article is to high-light examples demonstrating this point based on theoriginal studies presented in this volume.

Another aim of this article is to introduce a new classi-fication of the platyrrhines that is designed to betterreflect advances made recently, since the first morphol-ogy- and cladistics-based overhauls reshaped our think-ing in the 1970s and 1980s. I draw particular attentionto this point. The need for a revised classification is notbased now on new phylogenetic hypotheses. Rather, itreflects major increases in our awareness of taxonomicdiversity. There are more than three times as many fos-sil genera known today than in 1977, for example, andnearly three times as many fossil genera as there areliving genera. Up until that point, classifications neces-sarily had to be anchored in information based on whatanimals exist today. However, new panoramas of platyr-rhine biodiversity are being brought to light. An exam-ple. For one of the major modern NWM clades, the Sakisand Uakaris (Pithecia Chiropotes, and Cacajao), theaddition of fossils has totally altered our picture of biodi-versity and the ecological role this group has played inplatyrrhine evolution (Rosenberger, 2002; Rosenbergeret al., 2009). Once seen as a backwater ensemble of lim-ited diversity and a nonfactor in terms of broader classi-fication schemes, the pitheciins now appears to be aremnant of a central ecophylogenetic force in the historyof the radiation. Like others who have recently advo-cated for the position (see below), I now regard thisadaptive array as a separate family, Pitheciidae. How-ever, this classification also comes with a caveat, for itcan only be provisional. With many more taxa now inneed of detailed study, the phylogenetic interrelation-ships of many fossil platyrrhines are still not wellunderstood.

Structurally, the article is presented in three sections.The first provides a brief background to platyrrhine tax-onomy and systematics. The second introduces the vol-

ume’s papers in a discussion of their new contributionsto platyrrhine evolution in the context of an evolutionarymorphology perspective. The third section is a targetedsummary and critique of the algorithmic approach as ithas been applied in morphology-based studies of platyr-rhine cladistics.

PART I. BACKGROUND TO PLATYRRHINESYSTEMATICS

Rosenberger (2002) argued that 1977 dated a para-digm shift in platyrrhinology. With the publication ofHershkovitz’s (1977) massive volume on callitrichines(Figs. 1,2), which also included many companion studiesof anatomical systems and their evolution among platyr-rhines as a whole, the nonfunctional (descriptive) ana-tomical perspective which dominated taxonomic thoughtfor decades came to a close, along with the ScalaNaturae-like (gradistic) model of evolution. During theperiod prior, little progress was made either in classifica-tion or phylogenetics since about the 1920s, when schol-ars such as Pocock (1917, 1920, 1925) and Gregory(1922)—evolutionary morphologists in approach—madeimportant contributions identifying, classifying andexplaining ‘‘natural groups.’’ In his own approach,Hershkovitz used static keys to organize the classifica-tion system, rather than the monophyly principle ush-ered in via Hennig (1966) and cladistics.

To achieve monophyletic groupings required dynamicanalyses of characters, predicated on hypotheses ofhomology and polarity. In a philosophical shift, system-atics embraced transformational thinking, which wasconsistent with functional morphology, though then, asnow, applied rarely in studies of primate systematics. Itbecame important to explain how (in what way) andwhy features might have changed. These newapproaches and views were promoted, in different ways,in cladistic studies of platyrrhines by Ford (e.g., 1980,1986) and Rosenberger (e.g., 1977, 1979b, 1980). The ini-tial approach advocated by Ford was algorithmic (usingWagner tree parsimony). Rosenberger used characteranalysis (see Hecht and Edwards, 1977) infused with‘‘adaptational analysis’’ (see Bock, 1977; Szalay, 1977,1981; Delson et al., 1977; Szalay and Bock, 1991; Rose-nberger, 1992). Homologies and polarities were imputedand tested by the comparative method (i.e., phenetics,anatomical and behavioral correlation, in-group and out-group commonality, temporal precedence, development,functional-transformational models of adaptation andbehavior, etc.). I also weighed characters according tomy level of confidence in character analysis decisions,emphasizing the uniqueness of the structures involvedand their adaptive roles in the biological gestalt of thetaxon. A central methodological concept shared by Fordand Rosenberger was the reconstruction of hypotheticalancestral morphotypes, along with behavioral andadaptive inferences that could be attributed to thesehypotheses. Wood and Harrison (2011) have recentlyre-emphasized the importance of inferring morphotypesas part of the phylogeny reconstruction process.

Classifications illustrate the differences between thesepre- and post-Hennigian approaches (Table 1). Hershko-vitz (1977) used a large number of suprageneric catego-ries in his scheme, including five family-level (familyand subfamily) units devoted to eight genera of fossils

PLATYRRHINES AND SYSTEMATICS 1957

alone. A similarly large number of higher taxa wereused for the living genera. A classification as split asthis lacks phylogenetic and adaptive coherence and can-not compliment diversity. The first example of the sys-tem I have followed is a composite of formalclassifications (e.g., Rosenberger, 1981, 1992) and subse-quent taxonomic additions based on new fossils andseveral topical studies. It is a marked contrast to theHershkovitz model, organized around a monophyletictwo-family system that tiers other monophyletic groupsat lower ranks. The increased number of fossils addedto the Platyrrhini during this period is important, foraccommodating a large number of diverse new taxawould inevitably stress any system based initially onorganizing the living forms, as this one was. However,what is most striking is the multiplication of generarelated, or potentially related, to the pitheciins Pithecia,Chiropotes, and Cacajao. These three were often col-lected into a separate subfamily since the middle 1800s(Rosenberger, 1981), but with revised cladistic thinkingand the addition of fossils, this group has now swollen toabout 14 genera. As noted previously (Rosenberger,2002), such dramatic changes in knowledge has impor-tant implications for considering the role of this clade inthe ecological history of platyrrhines.

The second Rosenberger classification in Table 1 is arevision that accommodates the exceptional diversity ofthe ‘‘pitheciid’’ group by according it full family rank.Recognizing this third family is consistent with the sug-gestions of molecular systematists who have, since the1990s, promoted a three-family classification using thesame taxonomic concepts as I use here but with onesignificant difference in content (see Schneider andRosenberger, 1996). Much of the rest, however, main-tains the same structure, although newly discovered fos-sils have been added. The pitheciids continue to be achallenge systematically, and this new model will

Fig. 1. An ecophylogenetic model of platyrrhine evolution based on the living forms (modified fromRosenberger, 2000). The major axes of differentiation, body size, food, and locomotion are emphasized.See Table 1 for a more specific taxonomic breakdown and a classification that includes fossils.

Fig. 2. Saguinus oedipus (Cotton-topped tamarin), a small-bodied,clawed callitrichine. From Elliot, 1913.

1958 ROSENBERGER

undoubtedly benefit from further adjustments as the in-ternal affinities of larger groups (e.g., Tribe Pitheciini)become better known, or when intersubfamily relation-

ships are better understood. For atelids, I have electedto use only one subfamily and two tribes as the terminol-ogy is well established and there appears to be no

TABLE 1. A comparison of platyrrhine classifications

Hershkovitz (1977) Rosenberger (1981–2002) Rosenberger (present)

Family Callitrichidae Family Cebidae Family CebidaeCallithrix Subfamily Cebinae Subfamily CebinaeCebuella Cebus CebusSaguinus Saimiri SaimiriLeontopithecus yNeosaimiri yNeosaimiri

Family Callimiconidae yLaventiana yLaventianaCallimico yDolichocebus yDolichocebus

Family Homunculidae Subfamily Callitrichinae yKillikaikeyHomuculus Tribe Callimiconini yAcrecebusyDolichocebus Callimico Subfamily Callitrichinae

Family Cebidae yMohanamico Tribe CallimiconiniSubfamily Cebinae Tribe Sagunini Callimico

Cebus Saguinus yMohanamicoSubfamily Saimiriinae Tribe Callitrichini Tribe Saguinini

Saimiri Callithrix SaguinusyNeosaimiri Cebuella Tribe Callitrichini

Subfamily Aotinae Leontopithecus CallithrixAotus Subfamily Callitrichinae inc. sed. Cebuella

Subfamily Callicebinae yMicodon LeontopithecusCallicebus yPatasola Subfamily Callitrichinae inc. sed.

Subfamily Atelinae Family Cebidae inc. sed. yMicodonAteles yBranisella yPatasolaBrachyteles ySzalatavus Family Cebidae inc. sed.Lagothrix yChilecebus yBranisella

Subfamily Alouattinae Family Atelidae ySzalatavusAlouatta Subfamily Atelinae yChilecebus

Subfamily Pitheciinae Tribe Atelini Family AtelidaePithecia Ateles Subfamily AtelinaeChiropotes Brachyteles Tribe AteliniCacajao Lagothrix Ateles

Subfamily Tremacebinae yCaipora BrachytelesyTremacebus Tribe Alouattini Lagothrix

Subfamily Stirtoniinae Alouatta yCaiporayStirtonia yParalouatta Tribe Alouattini

Subfamily Cebupitheciinae yProtopithecus AlouattayCebupithecia Subamily Pitheciinae yStirtonia

Family Xenotrichidae Tribe Pitheciini yParalouattayXenothrix Pithecia yProtopithecus

Chropotes ySolimoeaSuborder inc. sed. Cacajao Family PitheciidaeFamily Branisellidae ySoriacebus Subamily Pitheciinae

yBranisella yProteropithecia Tribe PitheciiniyNuciruptor PitheciayCebupithecia Chropotes

Subfamily Pitheciinae inc. sed. CacajaoyLagonimico yProteropitheciayCarlocebus yNuciruptoryAntillothrix yCebupithecia

Tribe Homunculini Tribe SoriacebinaeyHomunculus ySoriacebusAotus (incl. A. dindensis) yMazzonicebusyTremacebus Subfamily HomunculinaeCallicebus yHomunculusyXenothrix Aotus (incl. A. dindensis)

yTremacebusCallicebusyMiocallicebusyXenothrixyAntillothrixyInsulacebus

Subfamily Homunculinae inc. sed.yCarlocebus

Family Pitheciidae inc. sed.yLagonimico


benefit to elevating the latter and adding additional tiersbelow at this time. As is evident, my philosophy aboutclassification is that it should be consistent with phylog-eny, first of all—and, hopefully, with an ecophylogeneticmodel—but that it also should be balanced as a taxo-nomic scheme relative to current concepts applied else-where. It should not in and of itself attempt to reflectoverwhelming adaptive departures detached from a clad-istic hypothesis by inflating ranks, as with the separa-tion of the extinct Jamaican genus Xenothrix at thefamily level, which has been advocated (e.g., Hershko-vitz, 1970, 1977).

It is pertinent in this regard to comment briefly onother schemes that have recently been published withregard to two taxonomic levels. Following several deca-des of taxonomic stability, we have entered a period of‘‘taxonomic inflation’’ (Issac et al., 2004), wherein popu-lations are being accorded full species status largelybecause new definitions or species concepts are beingapplied, less so because new knowledge has been gained.Issac et al. note that this phenomenon is more egre-giously evident in classifications of New World monkeysthan other primates. A similar effect is also evident atthe genus and family levels. At least three additionalgenera of living NWM are currently being recognizedbeyond the 16 identified by Hershkovitz (1977) andNapier (1976), the cornerstones of modern platyrrhineclassification. 1) Van Roosmalen and van Roosmalen(2003) differentially diagnosed Callibella (relative toCebuella) based on its slightly larger size and coat colorand pattern difference, a methodology that was univer-sally rejected for most of the 20th century. As I interpretthem, multivariate analyses of craniomandibular mor-phology also barely separates the three known specimensallocated to this form from Cebuella pygmaea (Aguiar andLacher, 2003), nor do univariate and multivariateanalyses of the postcranium (Ford and Davis, 2009). 2)Rylands et al. (2000) elevated Amazonian marmosets tothe genus level (Mico) essentially to avoid purported para-phyly of the classical concept of Callithrix, but there is nomorphological support for this view (see Hershkovitz,1977). 3) Groves (2001) advanced the notion thatLagothrix flavicauda should be elevated to the rank ofgenus (named Oreonax), but this was shown to be basedon a flawed parsimony analysis of poorly selected cranialtraits (Matthews and Rosenberger, 2008). At a higherlevel, this inflationary trend has emboldened Rylands andMittermeier (2009) to recognize five (!) platyrrhine fami-lies, with one designed exclusively for genus Aotus, andwithout considering fossils at all.

These taxonomic moves undermine the stability ofplatyrrhine classification that has been achieved over thelast 30 years based on knowledge of morphology,adaptation, molecules, and a consistent interpretation ofphylogenetics. Various recent articles review the cladisticmodels (e.g., Schneider et al., 2001; Schrago, 2007; Oster-holtz et al., 2009; Wildman, 2009; Kay et al., 2008; Rose-nberger et al., 2009). In my view, there remains oneinteresting and vexatious problem concerning the affin-ities of Aotus, for which morphology and molecules do notalign (e.g., Rosenberger and Tejedor, in press). One schoolplaces Aotus among pitheciids and the other among cebidssensu Rosenberger (1981). A few other discrepancies existin branching sequences when comparing morphologicaland molecular dendrograms, each important but less vital

than the Aotus paradox. Thus, the revised, non-Hershko-vitzian picture of platyrrhine systematics that began toemerge in the 1980s based on morphology, began to becorroborated by genetics in the 1990s. And the power ofthese adjustments has been demonstrated by the capacityof the new system to accommodate many new fossils with-out requiring still another major overhaul.

PART II. EVOLUTIONARY MORPHOLOGYAND PLATYRRHINE SYSTEMATICS

The global hypothesis that developed during the 1980sis an ecophylogenetic model of platyrrhine evolution (e.g.,Rosenberger, 1980, 1981, 2002; Rosenberger et al., 2009).It identifies four monophyletic groups, based on derivedcharacters and character complexes that are also func-tionally and adaptively important, whose homologies andpolarities were informed by functional-adaptive analysis,that is, evolutionary morphology. The three predominantecological niche parameters that initially characterizedthese adaptive patterns involved body size, diet, and loco-motion. A synopsis is depicted in Fig. 1.

The Size Factor: Body Size and Locomotion

Body size has always figured prominently in narra-tives of platyrrhine evolution because the group includesthe smallest living anthropoid, the 100 g Cebuella, and itsbiggest species, Brachyteles arachnoides, is roughly 100-fold larger, weighing about 10 k (Rosenberger, 1992; Fordand Davis, 1992; Peres, 1994). These species are alsoemblematic of two uniquely platyrrhine locomotoryorgans, digital claws (the only other exception amongeuprimates is the clawed aye-aye, Daubentonia), and pre-hensile tails, both foundational features of major NWMclades. Hershkovitz (1977) essentially saw the evolution ofsize in orthogenetic terms with the smallest platyrrhinesbeing the most primitive and the largest most derived.The idea that the miniature callitrichines are unusual orderived in being so small has been around for a long time(e.g., Pocock, 1917; Gregory, 1922), and a variety of argu-ments have been put forth in recent decades supportingthe hypothesis that they are secondarily dwarfed (e.g.,Ford, 1980; Leutenegger, 1980; Martin, 1992). Thesemodels tend to identify the special properties and combi-nations of features in the callitrichine bauplan as size-related in comparison with larger extant primates (e.g.,clawed digits, tricuspid molars, posterior tooth reduction,chimeric twinning, simplex uterus, large neonates). Manyof these features were explicitly considered in characteranalyses to test homologies, establish polarities, andexplain their co-occurrence as parts of an adaptive pack-age to facilitate occupation of a unique small-bodied nichein the platyrrhine radiation. Ford’s (1980) synthesiscontinues to be a useful statement in this regard.

The problem of callitrichine claws has held a particular fas-cination. More than 100 years ago, Wortman (1904:23, 24)said:

Whether the lack of opposability of the pollexand hallux [in callitrichines] is to be looked on as a degen-eration from a former more perfect condition of prehensil-ity of the extremities, or whether it represents a stage inthe process of acquirement of the opposability of thesedigits, cannot now be determined : : :

1960 ROSENBERGER

The habits of the marmosets, while as strictly and com-pletely arboreal as in any of the Primates, resemblethose of the squirrels more than those of the monkeysproper : : : the Negro Tamarin (Midas ureulus) confinesitself mostly to the larger branches, and is frequentlyseen passing up the perpendicular trunks, clinging tothe bark with its claws in a manner not dissimilar tothat of the squirrels. This method of climbing is doubt-less true of all the marmosets, and the lack of oppos-ability of the hallux and pollex is correlated with thepossession of sharp compressed claws instead of flat-tened nails.

Cartmill’s (1974) empirical demonstration of the claw’sbiomechanical value in the positional behavior of smallprimates cast new light on the matter to which Wortmanwas so attuned, and cleared the way for a new consensusto form around the hypothesis that claws and small bodysize are intertwined derived conditions. This allowedother pieces of the callitrichine puzzle to fall into place.The sensibility of this hypothesis is underscored by thefact that a morphocline of nail-to-claw shapes existsamong living platyrrhines, with many species exhibitingtransversely arched and prolonged nails on all digits butthe hallux, which is uniformly flat nailed.

With new observations and a sensitive quantitativemethod, Maiolino et al. (2011) reopen the riddle of pla-tyrrhine nails and claws. They show that groomingclaws, and a correlatively specialized terminal phalanxto affix them, exist on pedal digit II in at least two non-callitrichine genera, Aotus and Callicebus (Fig. 3)—something known to the literature but long since forgot-ten. This places platyrrhines, again, is a position topotentially inform us about the morphotype condition ofanthropoids which heretofore was assumed, perhaps

incorrectly, to have lost the grooming claw as a synapo-morphy: in preserving grooming claws, platyrrhines mayhave retained the ancestral anthropoid condition. Thecombined occurrence of nails and grooming claws on thefeet of platyrrhines, tarsiers, and living strepsirhinesalso raises a larger question of homology and transfor-mational process germane to primates more broadly, aswe mention. If these structures are phylogenetically con-tinuous with counterparts found in the other greatclades of euprimates, an additional question is raisedabout the actual homologies of the grooming claws. Didplatyrrhine grooming claws evolve from the originaleuprimate grooming claws, which could have arisendirectly as modified falculae without passing through aflat-nail character state?

The unique positional behavior of clawed callitrichinesdoes not particularly stand out, as one might expect, inthe Youlatos and Meldrum (2011) review of the locomo-tion of extant NWM and well as the fossil platyrrhines.This is the most up to date survey and evaluation of theplatyrrhine locomotor evolution produced during the last20 years. The authors present a multivariate analysis offrequency data describing locomotion as well as the over-all morphology of the astragalus. The astragalus cap-tures the diversity of platyrrhine ankle morphology andsorts fairly well along the lines of the major ecophyloge-netic divisions, though not without some interestingoverlaps and divergences. Most of the cebids form a clus-ter, which then links with Aotus and Callicebus—allspringers in the older terminology (e.g., Erikson, 1963).

However, the two most distinct sets in this aspect oftheir study are the large, quadrupedal, prehensile-tailed,and suspensory atelids and the quadrupeds Cacajao andChiropotes, all taxa that use pedal suspension and footreversal postures. It is rare for studies of positional

Fig. 3. The pitheciids Aotus vociferans (Spix’s night monkey), left, and Callicebus personatus (Maskedtit), right. Morphology indicates they are closely related; molecules suggest Aotus is more closely relatedto callitrichines and cebines. From Spix, 1823.


behavior to examine the transitions between the catego-ries to which we assign animals, but Youlatos andMeldrum use the concept of preadaptation to offer amodel pertinent to the evolution of tail-suspension. Theysuggest the manner of below-branch foot-hanging seenin some pitheciins (Fig. 4), which is accompanied by adraping of the tail (tail-bracing) over the supporting sub-strate, may represent a behavioral antecedent to the ate-line tail-hang. This presupposes an ancient behavioral/adaptive continuity between two sister-clades, pitheciids,and atelids, a notion also promoted by Rosenberger et al.(2011) in suggesting a transition between seed- and leaf-eating.

The positional behavior of Cebupithecia, an iconic fos-sil pitheciin from the middle Miocene of Colombia, is thesubject of Organ and Lemelin’s (2011) in-depth analysisof tail functional morphology. While it has been knownthat this species probably did not have a prehensile tail,Organ and Lemelin show the proximal tail region wasdesigned biomechanically to sustain relatively highbending and torsional forces, like prehensile- and semi-prehensile-tailed monkeys, but the distal end was differ-ent. They conclude that Cebupithecia, like one of theliving Saki species, Pithecia monachus, as well as Chiro-potes, was probably capable of tail-bracing postures.Organ and Lemelin also suggest this behavior may havebeen a primitive feature of the pitheciin clade. This is anintriguing idea because, as they say, there is no evidencefor hindfoot reversal in Cebupithecia. One other reason

why their argument intrigues is that it extends toMedrum’s (1998; Youlatos and Meldrum, 2011) preadap-tive hypothesis regarding the evolution of the atelid pre-hensile tail. Identifying the tail-bracing morphology andbehavior in a fossil pitheciin and pushing its originsback to the pitheciine morphotype may mean tail brac-ing was inherited from the last common ancestor sharedby pitheciids and atelids.

Returning to the matter of body size, bigness hasbecome a focal point of attention with the remarkablediscovery of two large Brazilian subfossils in the 1990s(see MacPhee and Horovitz, 2002). Their existence hasessentially doubled the established upper point limit ofNWM in terms of body mass. The ecological consequen-ces of this new datum has still to be explored, but occu-pying a class which places these platyrrhines squarelyinto a size-niche dimension that parallels Old Worldmonkeys for the first time may be revelatory. In fact, ithas even been argued that platyrrhines had alreadypeaked at an intrinsic size limit with Brachyteles (Peres,1994), the largest modern NWM. However, the two sub-fossils, Protopithecus and Caipora, are also atelines, andseem to be about twice as large (Hartwig and Cartelle,1996; Cartelle and Hartwig, 1996).

This inference is now strengthened by Halenar(2011a). She reassesses the body size estimates of Proto-pithecus using a variety of regression models and skele-tal elements, including, for the first time, goodbackground samples of platyrrhines to exert tight

Fig. 4. The pitheciids Chiropotes satanus israelita (Brown-backed bearded saki), left, Pithecia pitheciacapillimentosa (White-faced saki), right, are closely related seed-predators. From Spix, 1823.

1962 ROSENBERGER

phylogenetic control. Halenar finds that Protopithecuswas baboon-sized, likely weighing about 20 k. In arelated article (Halenar, 2011a), she evaluates the mor-phology of the distal humerus and proximal ulna usingthree-dimensional (3D) morphometrics. Here, Halenarfinds another unexpected result. Protopithecus, analouattin, while exhibiting limbs that are quite robustand a skull that resembles Howler monkeys in some un-usual features, has an elbow that resembles more closelythe hyperdynamic suspensory locomotors Ateles and Bra-chyteles than the quadrupedal climber Alouatta, a genussometimes described as sluggish. Halenar suggests anadjustment in the model of the ancestral ateline locomo-tor pattern (see Rosenberger and Strier, 1989), whichshe envisions as being more acrobatic. She also notesthat Protopithecus may have been fully capable of someground use. This compliments the interpretation of Mac-Phee and Meldrum (2006) with regard to the relatedCuban subfossil alouattin Paralouatta, which they thinkmay have been semiterrestrial.

One of primatology’s most arresting cases of parallelevolution, the independent development of prehensiletails in Cebus (Fig. 5) and the atelids (or semiprehensileand prehensile, respectively, depending on one’s pre-ferred terminology) is illuminated by new observationsmade by Organ et al. (2011). For generations, this simi-larity led zoologists to associate Cebus and atelids inclassifications, which in today’s frame would imply theyare monophyletic. Such a perspective posed a seriouschallenge to the post-Hershkovitzian phylogenetic

hypotheses that placed Cebus and Saimiri with callitri-chines as opposed to atelids or any other platyrrhines.Functionally and behaviorally oriented character analy-sis (Rosenberger, 1983), followed by a diverse series offield studies (see Garber, 2011) as well as morphologyprojects (e.g., Organ, 2010), showed that the essentialCebus-atelid similarities were best interpreted as analo-gies. Organ and coworkers now demonstrate this phe-nomenon at the histological level. In the Cebus tail,which is fully clothed in fur, the critical mechanicorecep-tor cells in the skin are suited mainly for heavy-pressuretouch. The glabrous patch of skin on ateline tails, in con-trast, are embedded with twice as many receptor types,and it is designed to sense both light and heavy pres-sure. They hypothesize that the Cebus system evolved tosupport sustained postural behaviors while the atelid’ssensitive arrangement evolved to support the delicatetiming dynamics of tail-infused locomotion.

The Food Factor: Fossils, Seeds, and More

The assessment of dietary adaptation has become avital special interest among primatologists—not surpris-ing as the mammalian fossil record is predominantlymade up of teeth—and platyrrhine groups are notablefor their fealty to distinct foods while also exhibiting ageneralist streak. As indicated above, the feeding guildsthat have been identified are also clades (Fig. 1), andthe particular foods emphasized by callitrichines,cebines, pitheciids, and atelids, and/or the way they

Fig. 5. The living predatory cebines include Saimiri sp. (Squirrel monkeys), left, and Cebus apella libid-onosus (Black-striped tufted capuchin), right, related monophyletically to the callitrichines. From Ferreira(1971) and Spix (1823), respectively.


access them via locomotion and posture, are at the eco-logical roots of their evolution.

In an effort to infer the diets of fossils using lowermolar morphology, Cooke (2011) developed a large data-set of laser surface-scanned 3D models and analyzedtheir patterns via 3D geometric morphometrics (3DGM).Her project emphasizes the extinct Caribbean forms butalso samples mainland Miocene species from Patagoniaand Colombia. An interesting facet of Cooke’s analysis isthe support it offers one of the models explaining selec-tion for molar adaptation, the critical function hypothe-sis, which was originally formulated in researchingplatyrrhine molar occlusion (Rosenberger and Kinzey,1976). Cooke found that the dietary extremes, generawhose foods—leaves, insects, hard or tough fruits, andseeds—would be most difficult to process mechanically,were offset from the other platyrrhines in shape, andmore easily classified according to diet using multivari-ate discriminant functions. Other platyrrhines, withmore mixed but largely frugivorous diets, did not sort aswell. There are also interesting faunal patterns inCooke’s data, although these need to be judged cau-tiously in view of the restricted samples used. Yet, theCaribbean and Patagonian primates seem not to presentidentifiable adaptations to insectivory, folivory, duroph-agy, or seed predation. These are major dietary pattersemblematic of the modern radiation, and they were alsoestablished among the middle Miocene La Venta prima-tes of Colombia, which is an ecological antecedent of thecurrent Amazonian primate fauna. While earlier in time

and separated far to the south geographically, the ab-sence of several feeding niches among Patagonian fossilsis no doubt partially due to sampling, but it also reinfor-ces the idea that the southern cone and the Caribbean,especially, supported ecologically distinctive primateassemblages (Rosenberger et al., 2009).

It is interesting that with Cooke’s (2011) system ofmeasurement, which is the most sophisticated attemptthus far to capture the coherent crown geometry of pla-tyrrhine molars, the phenetics, even with the classifica-tory power of 3DGM, does not consistently reflectphylogenetics at the suprageneric level. However, itsdoes appear to be sensitive to intrageneric and inter-generic relationships in some cases. For example, Cookeconfirms the unusual occlusal morphology and molarproportions of the Jamaican Xenothrix and finds it clus-ters closely with the new Hispaniolan fossil Insulacebusin shape space (see Cooke et al., 2011). She finds corrob-oration for congeneric status of modern Aotus with amiddle Miocene Laventan species, A. dindensis, long amatter of controversy (e.g., Kay, 1990; Rosenbergeret al., 1990). The sensitivity of the approach is alsoapparent in another difficult example. It is able see thatthe Caribbean forms exhibit interesting morphologicalsimilarities with the early Miocene Patagonian Soriace-bus (see Rosenberger, 2002), whose incisor, canine, andpremolar functional morphology suggests pitheciineaffinities (Rosenberger et al., 1990) despite oddly shapedmolars not exhibited among modern NWM or other fos-sils (but see Tejedor, 2005). This observation goes to the

Fig. 6. Alouatta caraya female and young (Black and gold hower), left, and Brachyteles arachnoidsright, are large-bodied, prehensile-tailed atelids. From Spix, 1823.

1964 ROSENBERGER

proposition that the Caribbean forms managed to enterthe Greater Antilles relatively early in the history ofthe crown group (see MacPhee and Horovitz, 2002;Rosenberger et al., 2011; Cooke et al., 2011).

These additional links between some Caribbean fossilsand the pitheciid NWM, which as noted above are turn-ing out to be a major, highly successful clade within theadaptive radiation, is a fascinating development as theliving forms include some of the most anatomicallyspecialized feeders in either Old and New World, thepitheciin seedeaters. As Norconk and Veres (2011) dis-cuss, other primates also rely on seed eating but noneseem to have so much of the mouth redesigned to fit thisfood type, which involves cracking open woody legumi-nous casings. By providing empirical measures of thetoughness and resistance of foods eaten, Noconk andVeres perform a welcome service to morphologists, whohave long recognized the explicit connection betweenselection for anatomical form and the physical propertiesof the environment with which structures interact.Another remarkable hard fruit adaptation occurs amongCebus monkeys, which Cooke (2011) shows is compara-ble to pitheciins in having flat molar teeth. However, inthis case, the most dramatic feeding behaviors meanpounding fruits with tools to extract what the monkeyswant to eat, and the molars are histologically reinforcedby thick enamel, unlike the pitheciins.

This contrast lends perspective on the recent proposalthat the early Miocene Argentine fossil Soriacebusevolved a seed-eating adaptation in parallel to the pithe-ciins (Kay et al., 2008). Kay and coworkers, using a par-simony analysis (see below), found that Soriacebus felloutside the monophyletic crown group of NWM, thusrequiring an alternative explanation for why theyshared such exclusively pitheciin synapomorphies astall, laterally compressed lower incisors, massive lowercanines with a triangular cross section, and a massivepyramidal p2—all part of the breaching and harvestingcomplex that is becoming fairly well understood in livingpitheciins. When their interpretation was challenged(Rosenberger, 2010b), Kay and Fleagle (2010) basicallyreplied that parallelism was common and to be expectedamong NWM. They proffered the example of graspingtails, which evolved twice. However, the lessons derivedfrom the dentition of Cebus illustrates just the converse,how parallelisms arise in different clades as asymmetricmorphological equivalents because form is conditionedby other adaptive compromises in each case, in additionto potentially massive differences in their genomes. Thisprinciple, an aspect of what Bock (e.g., 1977) called‘‘paradaptation,’’ emerges from functional morphology,but it is explicitly a phylogenetic hypothesis as well. Theverisimilitude of the Soriacebus and pitheciin anteriordental complex, and the deeply inflated jaws shown bythe fossil as well, another potentially correlated pithe-ciine or pitheciin synapomorphy, suggest both a com-monality of function as well as phylogenetic affinity: thisis the null hypothesis. It is corroborated by the paradap-tive mechanical parallels of the Cebus pattern. There,similar biological roles are evident—open the casing,remove the innards, crush it—but those behaviors areperformed by a system design constrained, one hypothe-sizes, by a unique heritage not shared with pitheciins.The convergently evolved methods of Cebus cannot enlistincisors or canines to extract the seeds like Soriacebus

would have and pitheciins actually do. And, the shallow,crushing jaws of Cebus are powered by a different mus-cular arrangement, without a comparable emphasis onthe masseters, unlike deep-jawed Soriacebus and thepitheciins.

Generally, seeds are viewed as a fundamentally dis-tinct dietary branch of frugivory, but Rosenberger et al.(2011) call attention to some suggestive similaritiesshared by seedeaters and leafeaters which may havebroad consequences pertaining to dietary evolution inplatyrrhines and other primates. We present a synthesisof anatomical, behavioral, adaptive and phylogenetic in-formation in asking why the two most folivorous platyr-rhines, Alouatta and Brachyteles (Fig. 6), do not conformin similar fashion to well developed models of primatefolivory and herbivory. Howlers tend to, but certainlynot the Muriki. The one important constant they shareinvolve teeth highly sensitive to selection for a leafydiet, that is, small incisors and large, crested molars. Wepropose that even in these ‘‘semifolivores,’’ who eat oneof the most nutritionally and digestively challenging ofprimate foods, other nondietary factors interact withselection in the animal’s gestalt and some may be morepowerful than the large costs of eating leaves, which iswhy parallelism has not driven them to be more similarto one another. Concerning the origins of their leaf-eat-ing habits, there may be a connection to seed-eating, forseed coats and pericarp tend to contain the same typesof chemical deterrents that bother folivores and requiregut specializations to deal with, secondary compounds.We propose that a level of seed-eating adaptations in thegut may have been a preadaptation that enabled ate-lines to shift from a frugivorous to a folivorous diet, andthat this kind of a transition could have happened sev-eral times in primate evolution. A likely parallel exam-ple involves the origins of colobines, but a similarpathway may also have been taken early in primate his-tory, as nontarsiiform euprimates emerged in the Paleo-gene, perhaps inheriting seed-eating proclivities from aplesiadapiform ancestry.

In all their diversity, seeds and leaves are not theleast of dietary challenges experienced by a platyrrhine,for the smallest species are exudativores. Marmosets usethe anterior teeth to gouge and wound trees in order tostimulate a flow of edible gum. Some of the unique fea-tures of the incisor and canine teeth associated withgum-gouging are well known (e.g., Rosenberger, 2010a;Hogg and Walker, 2011), but how this behavior influen-ces craniomandibular anatomy is a matter of discussion.Forsythe and Ford (2011) find that gouging marmosetsdiffer from nongouging tamarins in only three of 25metrical features thought to be biomechanically impor-tant. They relate the differences to maximizing gape anddissipating loads in the face.

These studies have implicitly corroborated the hypoth-esis that the suite of craniodental features long thoughtto be indicative of the primitiveness of callitrichines(e.g., Hershkovitz, 1977) are not continuous with traitsfound in early primates or primitive mammals but are,instead, novel biomechanical specializations. On theother hand, as the Miocene fossil record of platyrrhinesbecomes better known, examples of callitrichin-like fea-tures are beginning to show up, such as the anteriorlynarrow lower jaws and staggered incisors that appear tobe present in Homunculus and Branisella. Full tests of


the potential homologies of these similarities arerequired, and the biomechanical patterns evident at verydifferent levels of biological organization identified byHogg and Walker (2011) and Forsythe and Ford (2011)should offer valuable perspective in such analyses.

So, too, will be the sophisticated functional morpholog-ical research reviewed by Vinyard et al. (2011), a stun-ning range of elegant lab work—in vivo strain gaugeassessment of the cranium and jaw; EMG; morphomet-rics; muscle fiber architecture. Even telemetered EMGstudies of Howler monkeys feeding in the wild. Criticallyimportant is their finding that ‘‘monkeys’’ of the OldWorld and New World are not biomechanical equiva-lents: neither Cebus nor Macaca can be idealized as typi-cal for purposes of applying dynamic functionalinformation to resolve the attractive questions abouthow australopiths worked, for example. To their credit,Vinyard et al. also focus on matters that have been stud-ied but produced counterintuitive results which demandfurther probing and, possibly, revisions of basic assump-tions. One such example echoes the results of Forsytheand Ford (2011): absence in gum-gouging marmosets ofenhanced resistance to loading the mandibularsymphysis.

Correlative Factors: Brains, Guts, and Behavior

As noted, the model of platyrrhine evolution thatshaped the previous discussion (Fig. 1) is essentially anecological hypothesis attempting to explain the shape ofthe NWM radiation by the animals’ differentiation alongthe principle axes of body size, food, and locomotion.However, there are more layers to this construct. Bodysize, food, and locomotion meld into the notion of forag-ing behavior, and foraging strategies trigger considera-tions of group size, interpersonal behavior, perception,life histories, cognition, and so forth.

For example, NWM are known to rely more heavily onolfactory communication than Old World anthropoids, somuch that in some callitrichine species, pheromones aresuspected of playing an important role in suppressingovulation among nonreproducing females, which makessmell a central organizing element of mating and socialsystems. However, the role of the vomeronasal organ(VNO) in mediating sociosexual behavior is complex andmuch in need of experimental research. Smith et al.(2011) provide a taxonomically broad morphological per-spective on this with their microstructural and macro-structural analysis of the VNO in platyrrhines, whichthey relate, in part, to selection involving competitionfor mates. They suggest that relatively smaller (shorter)vomeronasal grooves, the bony trough in which theorgan sits, may be associated with monogamy, wheresexual selection may be relaxed. However, short groovesare also present in the polyandrous callitrichines, whichleads the authors to suggest that the functional divisionof vomeronasal and main olfactory systems may be lessdramatic than supposed. One suspects this may alsomean that polyandry in callitrichins evolved from aprimitive condition of monogamy, perhaps in concertwith the extra parental care necessary for rearing twins.Nevertheless, by showing that VNO size has an osseouscorrelate, Smith et al. provide a method for tracing it inthe fossil record as a way of gaining new glimpses of thebehavior of extinct species.

Also investigating the special senses, Muchlinski et al.(2011) test the hypothesis that female Cebus monkeys,who must naturally experience higher levels of parentalinvestment than males, ought to be choosier about thefood they eat and should therefore be more sensitivetasters. To get at this question, the authors look at sexdifferences in the histology of the fungiform papillae(FP) on the anterior tongues in a series of capuchins.Their hypothesis is corroborated in finding that femaleshave 50% more FPs per square centimeter than males,and larger surface areas of FPs as well, although bothsexes appear to have similar total counts of taste pores(the actual receptors) within the papillae. This is apotentially rich area of research with so many connec-tions to other evolutionary topics that it strains theimagination.

The relationships of dental growth with several lifehistory variables among the cebids, including brain sizeand body size, as well as ecological aspects, are exploredby Hogg and Walker (2011) in a tightly controlled phylo-genetic sample that varies in size by about 30 orders ofmagnitude and includes differing levels of encephaliza-tion. Theirs is the first extensive study of enamel growthrates in platyrrhines, and they use three different meas-ures of ontogenetic development. Hogg and Walker sug-gest that although brain mass strongly influences timingof dental eruption statistically, ecological factors, partic-ularly the age of foraging independence, have a strongerinteraction with enamel growth. These results are notentirely consistent with patterns found among strepsir-hine primates, which raises a variety of questionsrequiring additional research. The first next step mightbe to sample atelids and pitheciids, to see if the platyr-rhines as a group are distinct.

Hartwig et al. (2011) examine the evolution of brainsand guts in platyrrhines. First off, we present a newrobust series of raw measurements on brain weight (andbody weight) that encouragingly validate the measure-ments of a more than 40-year-old data set on brain mass(Stephan and Andy, 1964; Stephan et al., 1981) that hasbeen the primary resource for many subsequent studieson primates. We also show the good correspondencebetween skeletally based endocranial volume measuresand actual brain weights. Then we examine relativebrain size in platyrrhines, including all modern generabut Brachyteles. A new finding is that Chiropotes andCacajao, seed-predators, have elevated brain sizes,approaching and perhaps equaling the high encephaliza-tion quotient of Cebus, which is famously regarded asone of the most highly encephalized (and intelligent) pri-mates. In suggesting that NWM increased relative brainsize in at least three lineages independently, we reviewsome of the ecological and social factors that mayhave driven these patterns. We also find a complex rela-tionship between relative gut size and differentiationvis-a-vis brain size.

In summary, this selected series of examples relatingto the status of evolutionary morphology as a didacticmethod in the world of NWM, are proof positive of anenlarging, vital research agenda. There is little doubtthat the post-Hershkovitzian models of platyrrhineinterrelationships and adaptation have shaped the direc-tion of this body of work at several levels, which perhapshas added to the value of their end results. At the sametime, while there has been room to challenge these

1966 ROSENBERGER

ecophylogenetic models as new data accumulated andwere often analyzed by independent methods, the stud-ies reported here tend to reinforce the essential cladistichypotheses. Whether it is toes or tails or sources ofnutrition, ideas about the adaptation and evolution ofanatomical features and complexes and adaptive strat-egies are inevitably rationalized through the phyloge-netic perspective, information gained more often thannot in the absence of fossils. The research on platyr-rhines thus becomes an example of the reciprocity offunctional morphology and phylogeny reconstruction,and the potential power of joining the two as an opera-tional, evolutionary morphology approach whenappropriate.

[T]he cult of impressive technicalities or the cult ofprecision may get the better of us, and interfere withour search for clarity, simplicity, and truth.

Karl Popper (1983, p. 60)Realism and the Aim of Science1

PART III. EXPERIMENTS IN PARSIMONY

Kay and coworkers (e.g., Kay, 1990, 1994; Kay andMeldrum, 1997; Kay and Cozzuol, 2006; Kay et al.,2008) have been the most ardent users of parsimonyalgorithms (PAUP; e.g., Swofford, 2002) as a method toreconstruct the cladistics of NWM based on morphology.The effort makes their work an indispensable resourcefor evaluating the power of this approach in a controlledsituation—one lab, using a consistent and expandingdataset based mostly on craniodental features, appliedto a single adaptive radiation. It is noteworthy that thisapproach led Kay et al. (2008) to suggest that nearly allof the early middle Miocene platyrrhines from Patago-nia, which comprises a very important collection of fos-sils key to understanding platyrrhine evolution, belongto a pre-crown radiation of NWM. This idea addressescentral questions about the tempo and mode of platyr-rhine evolution: How old are the extant lineages andhow much have they changed over time? The model pro-posed by Kay et al. has been presented as an alternativeto the Long Lineage Hypothesis (e.g., Rosenberger,1979a; 2010b; Delson and Rosenberger, 1984), whichholds that some of these same Patagonian forms knownby interpretable material are actually early representa-tives of enduring extant clades, in a broad sense repre-senting the ancestral groups from which platyrrhinesevolved into the modern guilds as the Amazonian faunaassembled several millions of years later (Rosenbergeret al., 2009). More accurately, it addresses a temporalquestion: When do these long-term lineages, documentedby molecular systematists as well (e.g., Opazo et al.,2006; Schrago, 2006; Osterholtz, 2007; Schneider et al.,2001, Wildman, 2009), first appear in the fossil record?

Figure 7 presents summary cladograms of this body ofwork (e.g., Kay, 1990, 1994; Kay and Meldrum, 1997;Kay et al., 2008), chronologically. The dendrogramsappear as published (a, b, and c) or as extracts takenfrom larger trees (d and e). They were selected to high-

light the interpretations of Callicebus, one of the moreinteresting genera whose affinities have historicallybeen problematic (see Rosenberger, 1981; Ford, 1986;Rosenberger and Tejedor, in press). The trees come fromfour different studies using an expanding dataset involv-ing 15 to 199 craniodental characters, and as many as30 primate genera. For one study, I reproduce two differ-ent dendrograms, one (e) based purely on morphologyand another (d) that use a ‘‘molecular backbone’’

Fig. 7. Five whole or partial maximum parsimony cladograms fromseveral studies produced by Kay and coworkers (see text). (a) Kay(1990): 117 dental characters, 19 taxa [genera], Consistency index (CI)¼ 0.45. (b) Kay (1994) Lagonimico tree: 18 characters (15 dental), 9taxa [genera]; CI ¼ 0.65. (c) Kay and Meldrum (1997) Patasola ‘‘pre-ferred’’ cladogram, 55 dental characters, 11 taxa [10 genera], CI ¼0.54. Retention index (RI) ¼ 0.55. (d) Kay et al. craniodental tree con-strained by molecular backbone tree (Kay et al., 2008, Fig. 20), CI ¼0.34. RI ¼ 0.53, 199 characters craniodental. 30 taxa [genera]. (e) Kayet al. craniodental tree unconstrained by molecular backbone (Kayet al., 2008, Fig. 21); CI and RI not provided. Compare (a) and (e) for afuller picture of branch-by-branch discordance of strictly morphology-based parsimony results.

1Cited byGrant andKluge (2003).


approach to reconstruct morphology, which essentiallyinvolves mapping morphological character states ontothe topology of an existing molecular tree. The purposeof this particular application was to generate a list oftraits useful for placing fossils within a presumablyrobust cladogram.

The five trees reveal five different linkages for Callice-bus. The first example (Kay, 1990) should perhaps betreated as an anomaly—a starter tree?—for the out-standing cladistic elements that give it shape haveessentially been abandoned by Kay in all subsequentarticles (e.g., compare with Fig. 7d,e), save for the recog-nition (but not the interrelationships) of three monophy-letic groups consisting of callitrichines, pitheciins andatelines. For all intents and purposes, when stripped ofthe wayward interpretations of nonspecialists seeking toforce fit classifications, these three clusters of generahave been understood to be coherent units for severalgenerations (Rosenberger, 1981). However, in 1990, Kayinterpreted Callicebus as a basal member outside thethree clades, in a position below Cebus at the beginningsof the crown radiation. There is little question today,from morphology and molecules, that both Callicebusand Cebus are nested well within the Platyrrhini (e.g.,Schneider and Rosenberger, 1996; Opazo et al., 2006;Schrago, 2006; Osterholtz, 2007; Schneider et al., 2001,Wildman, 2009).

However, in each of the four other morphology studiesin the series Callicebus obtains different positions aswell, sharing an immediate node with Pithecia (b),Cebus (c) and Aotus (e), or shown as the sister-group topitheciins (d). There are many interesting aspects tothese results. (1) Aotus is nearly always in the mix,never being more than one step removed from the Calli-cebus twig or node. Aotus and Callicebus have long beenthought to be closely related (see Rosenberger, 1981). (2)In only one ‘‘unconstrained’’ case (b) is Callicebus tieddirectly to Pithecia, but not when the other pitheciins,Chiropotes and Cacajao, are included in the sample (dand e). Now, the close cladistic association of Callicebusand pitheciins is unanimously accepted by morphologistsand molecular systematists (e.g., Rosenberger, 1992;Schneider et al., 2001; Opazo et al., 2006; Schrago, 2006;Osterholtz, 2007; Rosenberger et al., 2009; Wildman,2009). (3) In two post-1990 iterations, Callicebus, Aotus,and Cebus form a exclusive monophyletic group (c ande). This contradicts the powerful cladistic linkagebetween Cebus and Saimiri, which is extended to encom-pass callitrichines as well by essentially all other studies(ibid.). (4) In none of the purely morphological iterations(a–c, e) is Aotus more closely related to cebines and calli-trichines than to any atelids. The former hypothesis hasreceived nearly universal corroboration from molecularsystematists since the middle 1990s (e.g., Schneideret al., 2001; Opazo et al., 2006; Schrago, 2006; Oster-holtz, 2007; Wildman, 2009), although I continue to beskeptical of this notion based on morphology, and alsobecause the molecular results seem soft. In a survey ofthem, Aotus is typically linked to cebids via polytomies,and its node receives uncommonly low and variable sta-tistical support (Rosenberger and Tejedor, in press).

These nonrepeated results concerning the interrela-tionships of Callicebus are major inconsistencies amongthe morphology-based parsimony tress. And, closeinspection reveals that the discrepancies are not limited

to one genus. They also involve Saimiri, Callimico,Leontopithecus, and so forth. While it is beyond thescope of this study to seek a specific explanation forerrant results, it seems clear that no particular outcomecan be said to be scientifically credible when it is notreplicated, especially by the same methods, in one lab,using a consistently similar body of information. Confi-dence in any one of the trees, or in particular nodes,could have been gained because each of these iterations,coindependent by their very nature thus highly likely torepeat, was slightly or more progressively tweaked bydesign, for example, with enlarging matrices of charac-ters and/or taxa but analyzed within the same experi-mental plan. However, this was not the case. Thestrongest common denominator among them in terms ofbroadly accepted and probably accurate results is foundin the monophyly of callitrichines, morphologically andbehaviorally one of the most derived platyrrhineclades—a finding not only expected but essentiallyrequired to avoid disqualification.

The final pair of cladograms was isolated from a largerstudy because it is instructive in another sense. One ofthe trees (e) is based exclusively on morphology and theother (d) is built from a molecular backbone. Again,prominent differences are seen with respect to Callice-bus, Aotus, Cebus, Saimiri, and so forth. For the firsttime in the series, the morphology (e) suggests Saimiriis not related directly to cebids (d), a conclusion that isstrongly contradicted by all other phylogenetic studiessince the post-Hershkovitz (1977) paradigm shift. How-ever, another aspect reveals the intrinsically limitedcladistic signal of the analysis itself. By mapping thecharacters onto the molecular backbone (d), Kay et al.(2008) were able to obtain a statistical assessment of the‘‘goodness of fit’’ between morphological characters andthe tree. The Consistency index (CI), which is designedto measure the amount of homoplasy, was 0.34, and theretention index (RI), which is meant to measure theamount of synapomorphy, was 0.53. Since plesiomorphyand homoplasy contain no credible cladistic informationby definition, the values obtained here indicate aboutone-third to one-half of the morphological charactersmapped onto the molecular tree can be regarded as clad-istically irrelevant noise. At most, only about half thetree can be reliably held, from this particular viewpoint,to be built on nodes supported by homologous derivedcraniodental characteristics.

While I do not assume the molecular tree has prece-dence over and is more likely to be more ‘‘right’’ thanthe morphological tree, the question is—What does thispoor fit mean? If the molecules are correct, is this repu-diation of morphology, or of the particular characters ormethods used in the morphological study? It should alsobe recalled that two of the prior, purely morphologicalstudies achieved much higher CI values, which is partlya function of having smaller matrices, but they alsoexhibited discordant branching sequences.

Kay et al.’s (2008) research strategy in this last study(d and e) has been critiqued at length (Rosenberger,2010b; but see Kay and Fleagle, 2010). Various theoreti-cal questions regarding specific sampling protocols andpossible artifacts relating to them have been pointedout. However, the backbone methodology is problematicfor other, deeper reason as well. It assumes, for example,a symmetrical correspondence between the patterns of

1968 ROSENBERGER

molecular and morphological evolution (e.g., Assis andRieppel, 2010). Moreover, Assis and Rieppel (2010:4)point out an epistemological dilemma associated withthe procedure that is inherent in the presumed primacyof the molecular evidence: ‘‘Using mapped, rather thattested, characters as synapomorphies cannot lend sup-port to the inference of monophyly, because such ‘‘syna-pomorphies’’ are empirically empty: they can never beshown to be wrong.’’ In other words, adhering to a back-bone methodology makes the results and all contingentanatomical hypotheses oblivious to tests usingmorphology.

Irrespective of this caveat, Kearney and Rieppel(2006) also note that the procedures used in algorithmicanalyses of morphology like the ones discussed here aresimilar to what Hull (1970) described as an ‘‘antitheoret-ical’’ method of ‘‘look, see, code, cluster’’ (pg. 31). Thiswas Hulls’ summary critique of early formulations of nu-merical taxonomy, a systematic approach that sought ahigh level of scientific objectivity but wound up beingrejected as sound phylogenetic methodology, partlybecause nothing is theory-free in science. While acknowl-edging that practitioners vary in how they apply themethod, its main contentions and philosophical positionsare quite similar to those embraced by advocates of algo-rithmic cladistics. To paraphrase Hull (1970:31), theirpurported value lies in: 1) being based on unweighted orequally weighted characteristics, with no a prioriweighting; 2) not being biased toward any scientifictheory or prior research on the topic, or to ideas appliedto character delineation, homologies, and taxonomicclustering.

Several other large points emerge from this assess-ment of the differing cladograms (Fig. 7). One has beendemonstrated many times before in various studies ofprimates (see references in Rosenberger, 2010b) andother taxa (e.g., Hillis, 1998), including meta-analysis ofdozens of morphological studies. The matrix configura-tion of a parsimony study has a profound influence onresults. In other words, when it comes to real worldapplications, the method may be inherently unstableand unreliable, because no sampling of biodiversity canavoid omitting taxa and lineages that have gone extinct,and the consequences of this loss of information areimponderable but also clearly variable. Thus, we mightassume every matrix is a skewed matrix. Wheeler(2005:71) makes this point in another way, which chal-lenges the heart of parsimony as a phylogenetic methodgrounded in homology, saying: ‘‘Given a single clado-gram, two features are homologous if their origin can betraced back to a specific transformation of a branch ofthat [Wheeler’s emphasis] cladogram, but the same pairof features may not be homologous on alternative clado-grams. Homology is entirely cladogram dependent : : : ’’

Wheeler’s (2005) extreme formulation is meant to elu-cidate a concept of homology but it is also another wayof saying parsimony analysis does not guarantee thatthe states attached to nodes are parts of a genetic con-tinuum of traits or a discrete genetic pathway ofdescent. It only promises to find the shortest tree possi-ble, that is, a tree with the fewest nodes, based on howselect characters are distributed among select taxa. Inother words, parsimony is an optimization method. It isnot a tool that probes information in search of geneticqualities. It works by antiseptically redistributing

character states among taxa in a manner deemed mostefficient by predetermined rules, which may have noth-ing to do with morphological evolution or the evolvabilityof a particular radiation. Along the way, it brands traitsas ‘‘synapomorphies,’’ not in the Hennigian sense ofshared derived genetic homologies but as commondenominators of taxonomic clusters. In reality, thesecommon denominators, which unless homologized andpolarized are nothing more than anatomical abstrac-tions, may be primitive, derived, or analogous. As the CIand RI values indicate, the final cladogram is thus sup-ported by an amalgam of cladistic and phenetic informa-tion. This is poignantly illustrated by the support valuesin the morphological cases presented above, and in thepoor performance of the craniodental characters whenmapped onto the molecular tree.

Does this disqualify parsimony as a tool for phylogenyreconstruction? It should for purists. However, not if weconsider parsimony a heuristic device that may bring usa step closer to discovering interrelationships, andaccept its many phylogenetic caveats at the same time.It does mean that a parsimony tree cannot be regardedas a true cladogram unless the input characters arestated to be (in theory) homologies, that is, unless it ispreceded by character analysis. Why does parsimonywork to the extent it does? Perhaps because a minimallyadequate number of traits coded at the outset are indeedlikely to be homologues. With this kind of ‘‘head start,’’in many cases even purely phenetic resemblances canlead to a natural, monophyletic grouping of related taxa.However, if the parsimony tree is a phenetic-cladistichybrid, it follows that only some of its nodes are trust-worthy. More research, using alternative means, wouldbe required to decide which ones deserve confidence.

Having seen some of the drawbacks of parsimonystudies, in the present context the question nowbecomes: What role might evolutionary morphology havein improving the design of these studies? What roles canfunctional morphology play, more broadly, in cladisticstudies? There are two operative levels in the process ofphylogeny reconstruction, the dataset (taxon and charac-ter identification and coding) and the analysis (treebuilding). Functional morphology provides a testing plat-form for cladistic hypotheses, at the levels of characterand clade. If one assumes adaptation is what drivesmuch of morphological evolution, we can test how sensi-ble cladistic hypotheses are by evaluating adaptive conti-nuity and/or discord among potentially related taxa. Forexample, the association of Callicebus with pitheciins isone of the newer features of platyrrhine cladistics, butfew researchers recognized any overlapping similaritiesbetween these groups until it was determined that theyshared in common the unusual pattern of tall, narrowincisors, in addition to some subtle details of upper inci-sor morphology (Rosenberger, 1979b). This synapomor-phic clue attained more cladistic power when it wasrecognized that the Callicebus condition was a goodmodel for an early adaptational ‘‘stage’’ in the evolutionof the hyperderived pitheciin incisor-canine battery (Kin-zey, 1992; Rosenberger, 1992). At the same time, some ofthe trenchant differences between Callicebus and pithe-ciins were spelled out as adaptive departures related tothe radically derived seed-eating habits evolved amongthe latter. This offered a rationale as to why close rela-tives looked less alike than propinquity of descent might


predict, and it provided a sound hypothesis for the polar-ities of hypothesized changes.

In a much simpler example, the power of the hypothesisthat Pithecia, Chiropotes, and Cacajao are monophyleticderives not only because one can generate a long list ofshared anatomical novelties, but because we understandtheir anterior teeth as being functionally integrated withthe cheek teeth—equally odd in shape—as an adaptivesystem geared to seed harvesting and processing. Thus,we attribute high ‘‘phyletic valence’’ to the characters.This set of features and analytical concepts has been usedto place fossils with very incomplete anatomies, such asthe early middle Miocene form Proteopithecia from Pata-gonia (Kay et al., 1998). In fact, the confidence placed inthese character hypotheses is so high that Proteopitheciais the only Patagonian platyrrhine recognized by Kayet al. (2008) as belonging to the crown NWM clade. Theother taxa, represented by characters without clear-cutfunctional explanations if any at all, he and his coworkersdesignate as stem platyrrhines. Interestingly, at the sametime, they deny that another Patagonian fossil, Soriace-bus, is also a pitheciine (see above) though it, too, exhibitsa number of the same high-weight pitheciin traits.

Obviously, for cases involving taxa characterized byextreme morphologies, functionally interpretable, onemight say homologies and polarities are simple enoughthat conventional character analysis can do the job andalgorithms are simply overkill. However, what about themany other cases? One argument invoked to support theuse of algorithms seems to be that parallelism and con-vergence is both so subtle and so common (see Kay andFleagle, 2010) that the matter should not be left to thealleged subjectivity of conventional morphological meth-ods. While the concern for parallelism is as old as Dar-winian systematics itself, the entrenched contemporaryview that parallelism is rampant can be attributed tometa-analyses of datasets which have attempted toempirically demonstrate the frequency of parallelism byexamining the CI and RI statistics in parsimony projects(e.g., Williams, 2007), as well as to the CI and RI valuesin individual studies such as the series under discussionhere. However, these comparative reports are fraughtwith difficulties, apart from the fact, recognized by Wil-liams and others, that the measures are not independentof matrix size and cannot thus be compared easily acrossstudies. More serious objections are that high levels ofhomoplasy appear to be 1) predicated on characters thatare themselves nonindependent, and 2) involve traitsthat may not actually be suitable for higher phylogenyassessments. This situation arises directly from the deci-sion to minimize assumptions about morphology whencrafting data matrices. For example, the practice of cod-ing serial homologues, for instance, hypoconulids onfirst, second and third molars, as three separate charac-ters is one such example appearing in the platyrrhineliterature (e.g., Rosenberger, 2010b) that violates severalprecepts: 1) hypoconulids vary at the population levelswithin some species and genera, making them difficultto code and untrustworthy at higher taxonomic levelsand across clades, where they are prone to functionalconvergence; 2) the theory of homology dictates thatsuch features strung together across adjacent morpho-logical units constitute one feature, not two or three.According to Sereno (2009:626), and others, ‘‘characterindependence and the mutual exclusivity of character

states’’ are the fundamental assumptions behind charac-ter state formulations.

The abundance of parallelisms detected by meta-anal-yses and individual parsimony studies is an issue, for allagree that it can become a significant nuisance factor inphylogenetic studies. As noted above, in the era of thesupermatrix—when a single molar can contribute adozen states or more—we are likely to find high degreesof parallelism because each one is an artificial subdivi-sion of an integrated functional complex under selection.The essential point is that excising traits from theiranatomical systems and treating them all separatelystrips them of their functional and genetic qualities [toSimpson (1944), relationships], making the effort toreconstruct phylogenetic interrelationships more chal-lenging because it also detaches the analytical processfrom evolutionary theory. The more and more functionalunits are subdivided, the more information is lost abouttheir genetic properties and their potential interactionwith selective forces; the more noise is added to the sys-tem; the higher the measure of homoplasy.

In summary, for platyrrhines, morphology-based parsi-mony studies do not have a good track record. The ‘‘sim-ple’’ clades can be easily retrieved, as they would likelybe on purely phenetic grounds, but problematic generawander about the trees in an inconsistent manner. Near-repeat studies do not replicate. Knowing these resultstend to fail the test of repeatability, they must be viewedskeptically. Meaning, some nodes in a large tree are likelyto be reasonable hypotheses of monophyly, but others arenot. If there is any rule of thumb that can be appliedhere, it must be that nodes supported by features thatmake sense functionally in the lives of the animal are theones deserving our attention. Those supported by volatile,atomized traits of limited independence and functionalvalue, are to be suspect. As a heuristic device, parsimonyanalyses have a place, but without testing independentlythe homologies of the characteristics involved a parsi-mony tree cannot be accepted as a truly cladistic model.However, the under-performance of these studies in theplatyrrhine domain should not be taken as evidence thatmorphology has failed. Or, that molecules are more reli-able as phylogenetic evidence. Or, that parallelism is stul-tifying. These difficulties lie with the methods we use, notwith the morphology itself.

CONCLUSIONS

The diverse set of morphological studies presentedhere, and fieldwork bearing on morphological questions,continues to expand the body of knowledge concerningthe adaptive radiation of NWM. Evolutionary morpho-logical studies of platyrrhine primates now cover a widerange of questions concerning the current ecological andbehavioral adaptations of New World monkeys. In bothgeneral and specific ways, they add confidence to themodels of platyrrhine phylogeny and adaptation thatbegan developing in the 1970s and early 1980s, whenplatyrrhines first became subject to cladistic studiesinformed by character analysis. They support the gen-eral model of the platyrrhines’ current status as an eco-phylogenetic array differentiated along the principleecological axes of body size, diet, and locomotion, andserve to align other factors such as social organization,brain size, perception, cognition, and so forth. Moreover,

1970 ROSENBERGER

they generate testable hypotheses about the adaptiveshifts that must have occurred as the platyrrhine pano-rama unfolded.

One reason for their explanatory power is that the think-ing behind these articles is closing the perceived gapbetween branches of morphology that are often held apartfrom one another, such as functional morphology and bio-mechanics, allometry, histology, systematics, and phylogenyreconstruction, while also addressing matters of ecology,behavior, and life history that articulate with anatomy. Theresult is a more expansive evolutionary morphology. Itsperspective provides a robust platform for generating andtesting hypotheses about platyrrhine evolution, how andwhy groups and characteristics changed over time, in addi-tion to serving morphology’s more conventional raisond’etre, to investigate how things are structured, how theyoperate and how to categorize anatomical diversity. Evolu-tionary morphology also provides a primary way to ration-alize false cladistic signals by exposing parallelisms asindependently evolved paradaptations.

The notion of evolutionary morphology ties phylogenyand adaptation together as a matter of method. Many ofthe hypotheses about the cladistic interrelationships ofplatyrrhines are made plausible because there arestrong adaptive hypotheses behind the characteristicsput forth to support the phylogenetic models, with func-tional morphology proving a deeper understanding ofthe connections between form and biological role, andbecause the groups so identified occupy an understand-able place in nature. Many of the homologies and polar-ities proposed as cladistic evidence in the first placewere identified as being of potential value because oftheir evident adaptive significance.

Molecular systematics offers an independent empiricaltest of morphology-based phylogenetic results, a directacid test of the hypothesized relationships and an indi-rect check of the evidence behind them. Overall, there isbroad concordance between molecular and evolutionarymorphological models, although some question marks—areas ripe for further research—still remain. What do Imean by broad concordance? Four major clades havebeen identified, and most workers see them as compris-ing two basic divisions of the radiation. Two of the fourare unanimously agreed to be sister-taxa. There is con-sensus on how the major subclades within these groupssort out. Many of the links between pairs and triplets ofgenera are agreed on. That the molecules and morphol-ogy are also in sync in terms of the time scale inferredfor the origins of these clades adds confidence andmutual corroboration of the phylogenetic outlines.

The main morphological method competing with thecharacter analysis approach to phylogeny reconstructionover the past 20 years has been parsimony-based cladis-tics. These projects suffer from inconsistency, as shownby following the results obtained for the genus Callice-bus in a series of related craniodental studies from thesame lab. Cladistically, differences in the placement ofCallicebus may be as small as one-node shifts or one-branch swaps, but they also get quite large: a cladisticposition nested deeply inside versus one at the very edgeof crown platyrrhines. In a taxonomic sense, the parsi-mony studies have aligned Callicebus with at least twodifferent families. A number of other genera (e.g., Sai-miri, Cebus, Aotus, Leontopithecus, and Saguinus) arealso variably placed, nontrivially, in these assessments.

The NWM parsimony studies have also been shownempirically to have a low cladistic yield at the characterlevel, that is, relatively few true synapomorphies, for thetrees tend to exhibit low levels of homology. The reasonsfor this are not easy to determine, but various critiquesof how the method has been used in primate studies,including platyrrhines, have pointed out a preponder-ance of issues. These could plague any phylogeneticstudy but they are more acute in rigid machine-basedtreatments. Among them: disparities in taxonomic sam-ples used; a reliance on population-level traits for higherphylogeny studies; matrices biased by taxon and charac-ter choices; excessive use of nonindependent, correlated,and redundant characters; and, especially acute for stud-ies involving fossils, matrices skewed by materiallimitations.

This relatively poor performance calls out for caution,especially when this approach is used to address, withina constrained taxonomic framework, the phylogenetics ofindividual, poorly known fossils. Without a complimen-tary assessment of the evidence by character analysis orevolutionary morphology, which has been shown to havea good track record, these projects are inevitably handi-capped by compromised samples: an overabundance ofmissing biological information and an overatomizedanatomy. Such is the case for several middle Miocenefossils of the circum-Amazonian basin. Examples areLagonimico (said to be a giant tamarin, but more likelya pitheciid) and Solimoea (said to be a primitive atelinbut more likely an aloauttin). Similarly, questionableresults have been obtained in assessing a whole group ofolder fossils from Patagonia. Based on a very unevenlypreserved information base, and using a molecular back-bone approach in an effort to extract cladistically inform-ative anatomical traits from the whole panoply ofmodern platyrrhine craniodental diversity, all the fossilswere determined to be outside the crown clade. In thiscase, the resemblances between Tremacebus and Aotus,and Soriacebus and pitheciins, of features shown to bepowerful ecophylogenetic indicators by evolutionary mor-phology, were rejected with little or no consideration,even though the trees’ homoplasy indices were high.

These examples call into question the trustworthinessof morphology-based parsimony as a tool for cladistics,certainly as the ultimate arbiter. If platyrrhines aretaken as an example, it has yet to be shown that parsi-mony-based methods, which tend to shun attaching evo-lutionary interpretations of homology, polarity, andadaptation as evidence inputs, have greater resolvingpower than a morphology which concentrates on fewercharacters and is laden with interconnected evolutionaryhypotheses. It has yet to be shown that anatomical infor-mation purged of its structural-functional context, thatis, evidence of its genetic covariation and inheritance,outperforms methods that fuse morphology, phylogeny,and adaptation. Or, that sheer quantity of information,including characters and taxa devoid of evolutionarycontext, redundancies, or those immaterial to thequestion at hand because they are autapomorphic forother groups, yields more robust results than characteranalyses that fuse morphology and adaptation within aproperly constructed phylogenetic framework.

As some have noted, the philosophical basis of algo-rithmic, morphology-based cladistics is akin to the ‘‘look,see, code, cluster’’ foundations of the now defunct school


of numerical taxonomy. In other words, there is ampleroom for improvement. On the other hand, the rich evo-lutionary morphological tradition that is taking root instudies of the platyrrhines, like those presented herein,suggests progress could be made even by inserting onlyone additional term to the protocol: look, see, interpret,code, and cluster.

ACKNOWLEDGEMENTS

For the access to specimens and resources, the authoris deeply indebted to many institutions but none morethan the American Museum of Natural History. Forencouragement and assistance in developing this specialvolume for the Anatomical Record, Associate Editor JeffLaitman stands out above all. Also, to Rosalie McFar-lane, Editorial Assistant, the author and coworkers owemuch. Special thanks all the authors and externalreviewers who contributed their knowledge, time, andskills, working under tight deadlines, to make thisendeavor possible. And thanks to collaborator MarceloTejedor for sharing freely his insights on platyrrhineevolution, which improved this manuscript.

LITERATURE CITED

Aguiar JM, Lacher TE, Jr. 2003. On the morphological distinctive-ness of Callithrix humilis van Roosmalen et al., 1998. NeotropPrim 11:11–17.

Assis LCS, Rieppel O. 2010. Are monophyly and synapomorphy thesame or different? Revisiting the role of morphology in phyloge-netics. Cladistics 26:1–9.

Bock WJ. 1977. Foundations and method of evolutionary classifica-tion. In: Hecht MK, Goody PC, Hecht BM, editors. Major patternsin vertebrate evolution. New York: Plenum Press. p 851–895.

Cartmill M. 1974. Pads and claws in arboreal locomotion. In:Jenkins FA, editor. Primate locomotion. New York: AcademicPress. p 45–83.

Cartelle C, Hartwig WC. 1996. A new extinct primate among thePleistocene megafauna of Bahia, Brazil. Proc Natl Acad Sci93:6405–6409.

Cooke SB. 2011. Paleodiet of extinct platyrrhines with emphasis onthe Caribbean forms: three-dimensional geometric morphometricsof mandibular second molars. Anat Rec 294:2073–2091.

Cooke SB, Rosenberger AL, Turvey S. 2011. An extinct monkeyfrom Haiti and the origins of the Greater Antillean primates.PNAS 108:2699–2704.

Delson E, Eldredge N, Tattersall I. 1977. Reconstruction of hominidphylogeny: a testable framework based on cladistic analysis.J Hum Evol 6:262–278.

Delson E, Rosenberger AL. 1984. Are there any anthropoid primateliving fossils? In: Eldredge N, Stanley SM, editors. Living fossils.New York: Springer Verlag. p 50–61.

Elliot, DG. 1913. A review of the primates. Monograph No. 1 Vols.1, 2 and 3. New York: American Museum of Natural History.

Erikson GE. 1963. Brachiation in New World monkeys and inanthropoid apes. Symp Zool Soc Lond 10:135–164.

Ford. SR. 1980. A systematic revision of the Platyrrhini based onfeatures of the postcranium. Ph.D. Dissertation, University ofPittsburgh.

Ford SM. 1980. Callitrichids as phyletic dwarfs, and the place ofthe Callitrichidae in Platyrrhini. Primates 21:31–43.

Ford SM. 1986. Systematics of New World monkeys. In: SwindlerD, editor. Comparative primate biology. Vol. 1: Systematics,evolution, and anatomy. New York: Alan R. Liss. p 73–135.

Ford SM, Davis LC. 2009. Marmoset postcrania and thee skeletonof the Dwarf Marmoset, Callibella humilis. In: Ford SM, PorterSM, Davis LC, editors. The smallest anthropoids: the Marmoset/Callimico radiation. New York: Springer. p 411–450.

Forsythe EC, Ford SM. 2011. Craniofacial adaptations to tree-goug-ing among marmosets. Anat Rec 294:2131–2139.

Garber PA. 2011. Primate locomotor positional behavior and ecology.In: Campbell CJ, Fuentes A, MacKinnon KC, Bearder SK, StumpfRM, editors. Primates in perspective, 2nd ed. Oxford: Oxford Uni-versity Press, p 548–563.

Grant T, Kluge A. 2003. Data exploration in phylogenetic inference:scientific, heuristic, or neither. Cladistics 19:379–418.

Gregory WK. 1922. The origin and evolution of the human denti-tion. Baltimore: Williams & Wilkins.

Groves C. 2001. Primate taxonomy. Washington DC: SmithsonianInstitution Press.

Halenar LB. 2011a. Reconstructing the locomotor repertoire ofProtopithecus brasiliensis I: body size. Anat Rec 294:2024–2047.

Halenar LB. 2011b. Reconstructing the locomotor repertoire ofProtopithecus brasiliensis. II. Forelimb morphology. Anat Rec294:2048–2063.

Hartwig WC, Cartelle C. 1996. A complete skeleton of the giantsouth American primate Protopithecus. Nature 381:307–311.

Hartwig W, Rosenberger AL, Norconk MA, Young Owl M. 2011. Rel-ative brain size, gut size, and evolution in New World monkeys.Anat Rec 294:2207–2221.

Hecht MK, Edwards JL. 1977. The methodology of phylogeneticinferences above the species level. In: Hecht MK, Goody PC,Hecht BM, editors. Major patterns in vertebrate evolution. NewYork: Plenum Press. p 3–51.

Hennig W. 1966. Phylogenetic systematics. Urbana: University ofIllinois Press.

Hershkovitz P. 1970. Notes on tertiary platyrrhine monkeys anddescription of a new genus from the late Miocene of Colombia.Folia primatol 12:1–37.

Hershkovitz P. 1977. Living New World monkeys (Platyrrhini) with anintroduction to primates, Vol 1. Chicago: University of Chicago Press.

Hillis, DM. 1998. Taxonomic sampling, phylogenetic accuracy, andinvestigator bias. Syst Biol 47:3–8.

Hogg RT, Walker RS. 2011. Life-history correlates of enamel micro-structure in Cebidae (Platyrrhini, Primates). Anat Rec 294:2193–2206.

Hull DL. 1970. Contemporary systematic philosophies. Ann RevEcol Syst 1:19–54.

Isaac NJB, Mallet J, Mace GM. 2004. Taxonomic inflation: its influ-ence on macroecology and conservation. TREE 19:464–469.

Kay RF. 1981. The nut-crackers—a new theory of the adaptations ofthe Ramapithecinae. Am J Phys Anthrop 55:141–151.

Kay, RF. 1994. ‘‘Giant’’ Tamarin from the Miocene of Colombia. AmJ Phys Anthropol 95:333–353.

Kay RF. 1990. The phyletic relationships of extant and extinct fossilPitheciinae (Platyrrhini, Anthropoidea). J Hum Evol 19:175–208.

Kay RF, Cozzuol MA. 2006. New platyrrhine monkeys from the soli-moes formation (late Miocene, Acre State, Brazil). J Hum Evol50:673–686.

Kay RF, Fleagle JG. 2010. Stem taxa, homoplasy, long lineages, andthe phylogenetic position of Dolichocebus. J Hum Evol 59:218–222.

Kay RF, Fleagle JG, Mitchell TRT, Colbert M, Bown T, Powers DW.2008. The anatomy of Dolichocebus gaimanensis, a stem platyr-rhine monkey from Argentina. J Hum Evol 54:323–382.

Kay RF, Johnson D, Meldrum DJ. 1998. A new pitheciin primatefrom the middle Miocene of Argentina. Am J Primatol 45: 317–336.

Kay RF, Meldrum DJ. 1997. A new small platyrrhine and the phy-letic position of Callitrichinae. In: Kay RF, Madden RH, CifelliRL, Flynn JJ, editors. Vertebrate paleontology in the neotropics.Washington DC: Smithsonian Institution Press. p 435–458.

Kearney M, Rieppel O. 2006. Rejecting ‘‘the given’’ in systematics.Cladistics 22:369–377.

Kinzey WG. 1992. Dietary adaptations in the Pitheciinae. Am JPhys Anthropol 88:499–514.

Leutenegger W. 1980. Monogamy in callitrichids: a consequence ofphyletic dwarfism? Int J Primatol 1:95–98.

MacPhee RDE, Horovitz I. 2002. Extinct Quaternary platyrrhinesof the Greater Antilles and Brazil. In: Hartwig WC, editor. Theprimate fossil record. Cambridge: Cambridge University Press.p 189–200.

1972 ROSENBERGER

MacPhee RDE, Meldrum J. 2006. Postcranial remains of the extinctmonkeys of the Greater Antilles (Platyrrhini, Callicebinae, Xeno-trichini), with a consideration of semiterrestriality in Paralouatta.Am Mus Novitat 3516:1–65.

Maiolino S, Boyer DM, Rosenberger A. 2011. Morphological corre-lates of the grooming claw in distal phalanges of platyrrhines andother primates: a preliminary study. Anat Rec 294:1975–1990.

Martin RD. 1992. Goeldi and the dwarfs: the evolutionary biology ofthe small New World monkeys. J Hum Evol 22:367–393.

Matthews LJ, Rosenberger AL. 2008. Taxonomic status of the Yellowtailed Woolly Monkey, Lagothrix flavicauda—an object lesson inthe implementation of PAUP*. Am J Phys Anthropol 137:245–255.

Meldrum DJ. 1998. Tail-assisted hindlimb suspension as a transi-tional behavior in the evolution of the platyrrhine prehensile tail.In: Strasser E, Fleagle J, Rosenberger A, McHenry H, editors.Primate locomotion: recent advances. New York: Plenum Press.p 145–156.

Muchlinski MN, Docherty BA, Alport LJ, Burrows AM, Smith TD,Paesani SM. 2011. Behavioral and ecological consequences of sex-based differences in gustatory anatomy in Cebus apella. Anat Rec294:2179–2192.

Napier PH, 1976. Catalogue of primates in the British Museum(Natural History). Part I. Families Callitrichidae and Cebidae.London: British Museum (Natural History).

Norconk MA, Veres M. 2011. Physical properties of fruit and seedsingested by primate seed predators with emphasis on Sakis andBearded Sakis. Anat Rec 294:2092–2111.

Opazo JC, Wildman DE, Prychitko T, Johnson RM, Goodman M.2006. Phylogenetic relationships and divergence times amongNew World monkeys (Platyrrhini, Primates). Mol Phylog Evol40:274–280.

Organ JM. 2010. Structure and function of platyrrhine caudalvertebrae. Anat Rec 293:730–745.

Organ JM, Lemelin P. 2011. Tail architecture and function of Cebu-pithecia sarmientoi, a Middle Miocene platyrrhine from La Venta,Colombia. Anat Rec 294:2013–2023.

Organ JM, Muchlinski MN, Deane AS. 2011. Mechanoreceptivity ofprehensile tail skin varies between ateline and cebine primates.Anat Rec 294:2064–2072.

Osterholtz M, Lutz W, Roos C. 2009. Retropositional events consoli-date the branching order among New World monkey genera. MolPhyl Evol 5:507–513.

Peres CA. 1994. Which are the largest New World monkeys. J HumEvol 26:245–9.

Pocock RI. 1917. The genera of the Hapalidae. Ann Mag Nat Histser 8. 20:247–258.

Pocock RI. 1920. On the external characters of the South Americanmonkeys. Proc Zool Soc Lond 1920:9–113.

Pocock RI. 1925. Additoinal notes of the external charavaters ofsome platyrhrien monkeys. Proc Zool Soc Lond 1925:27–47.

Popper K. 1983. Realism and the aim of science. New York: Rout-ledge. (Cited by Grant T, Kluge A. 2003.)

Rosenberger AL. 1977. Xenothrix and ceboid phylogeny. J Hum Evol6:541–561.

Rosenberger AL. 1979a. Cranial anatomy and implications of Doli-chocebus gaimanensis, a late Oligocene ceboid primate. Nature279:416–418.

Rosenberger AL. 1979b. Phylogeny, Evolution and Classification ofNew World Monkeys (Platyrrhini, Primates). Ph.D. Dissertaton,City University of New York.

Rosenberger AL. 1980. Gradistic views and adaptive radiation ofplatyrrhine primates. Z Morphol Anthropol 71:157–163.

Rosenberger AL. 1981. Systematics: the higher taxa. In: Coimbra-Filho AF, Mittermeier RA, editors. Ecology and behavior of neo-tropical primates, Vol 1. Rio de Janeiro: Academia Brasileira deCiencias. p 9–28.

Rosenberger AL. 1983. Systematics: the higher taxa. In: Coimbra-Filho AF, Mittermeier RA, editors. Ecology and behavior ofneotropical primates. Vol 1. Rio de Janeiro: Academia Brasilia deCiencias, p. 9–27.

Rosenberger AL. 1992. Evolution of feeding niches in New Worldmonkeys. Am J Phys Anthropol 88:525–562.

Rosenberger AL. 2000. Platyrrhini. In: Delson E, Tattersall I, VanCouvering JA, Brooks AS, editors. Encyclopedia of human evolu-tion and prehistory. New York: Garland Publishing. p 565–567.

Rosenberger AL. 2002. Platyrrhine paleontology and systematics:The paradigm shifts. In: Hartwig WC, editor. The primate fossilrecord. Cambridge: Cambridge University Press. p 151–159.

Rosenberger AL. 2010a. Adaptive profile versus adaptive specializa-tion: fossils and gumivory in primate evolution. In: A. BurrowsAM, Nash L, editors. The evolution of exudativory in primates.New York: Springer. p 273–296.

Rosenberger AL. 2010b. Platyrrhines, PAUP, parallelism, and thelong lineage hypothesis: a reply to Kay et al. (2008). J Hum Evol59:214–217.

Rosenberger AL, Halenar L, Cooke SB. 2011. The making of platyr-rhine semifolivores: models for the evolution of folivory in prima-tes. Anat Rec 294:2112–2130.

Rosenberger AL, Halenar LB, Cooke SB, Tejedor MF, Hartwig WC.(in press) Fossil alouattins and the origins of Alouatta: cranioden-tal diversity and interrelationships. J Hum Evol.

Rosenberger AL, Kinzey WG. 1976. Functional patterns of molarocclusion in platyrrhine primates. Am J Phys Anthropol 45:281–298.

Rosenberger AL, Setoguchi T, Shigehara N. 1990. The fossil recordof callitrichine primates. J Hum Evol 19:209–236.

Rosenberger AL, Strier KB. 1989. Adaptive radiation of the atelineprimates. J Hum Evol 8:717–750.

Rosenberger AL, Tejedor MF. (in press). The misbegotten: long line-ages, long branches and the interrelationships of Aotus, Callicebusand the saki-uakaris. In: Barnett AL, Veiga S, Ferrari S, NorconkMN, editors. Evolutionary biology and conservation of Titis, Sakisand Uacaris. Cambridge, Cambridge University Press.

Rosenberger AL, Tejedor MF, Cooke SB, Halenar L, Pekkar S. 2009.Platyrrhine ecophylogenetics, past and present. In: Garber P,Estrada A, Bicca-Marques JC, Heymann EW, Strier KB, editors.South American primates: comparative perspectives in the studyof behavior, ecology and conservation. New York: Springer. p 69–113.

Rylands AB, Mittermeier RA. 2009. The diversity of the New Worldprimates (Platyrrhini). In: Garber PA, Estrada A, Bicca-MarquesJC, Heymann EW, Strier KB, editors. South American primates:comparative perspectives in the study of behavior, ecology, andconservation. New York: Springer. p 23–54.

Rylands AB, Schneider H, Languuth A, Mittermeier RA, Groves C,Rodriguez-Luna E. 2000. An assessment of the diversity of NewWorld primates. Neotrop Prim 8:61–93.

Schneider H, Rosenberger AL. 1996. Molecules, morphology andplatyrrhine systematics. In: Norconk MA, Rosenberger AL,Garber PA, editors. Adaptive radiations of neotropical primates.New York: Plenum Press. p 3–19.

Schneider H, Canavez FC, Sampaio I, Moreira MA, Tagliaro CH,Seuanez HN. 2001. Can molecular data place each neotropicalmonkey in its own branch? Chromosoma 109:515–523.

Schrago CG. 2007. On the time scale of the New World primatediversification. Am J Phys Anthropol 132:344–354.

Sereno PC. 2009. Comparative cladistics. Cladistics 25:624–659.Simpson GG. 1944. Tempo and mode in evolution. New York: Co-lumbia University Press.

Smith TD, Garrett EC, Bhatnagar KP, Bonar CJ, Breuning AE, Den-nis JC, Kinznger JH, Johnson EW, Morrison EE. 2011. The vomero-nasal organ of New World monkeys (Platyrrhini). Anat Rec294:2158–2178.

de Spix J. 1823. Simiarum et vespertiliarum brasiliensis speciesnovae; or, Histoire naturelle des especes nouvelles de singes et dechauve-souris observees et recuellies pendant de vayage dansl’interieur di Brasil. Monaco.

Stephan H, Andy OJ. 1964. Quantitative comparison of brain struc-tures from insectivores to primates. Am Zool 4:59–74.

Stephan H, Frahm H, Baron G. 1981. New and revised data on vol-umes of brain structures in insectivores and primates. Folia Pri-matol 35:1–29.

Swofford DL. 2002. PAUP. Phylogenetic analysis using parsimony.Version 4.0b10 (Altivec). Sunderland, Sinauer Associates.


Szalay FS. 1977. Constructing primate phylogenies: a search fortestable hypothesis with maximum empirical content. J HumEvol 6:3–18.

Szalay FS. 1981. Functional analysis and the practice of the phylo-genetic method as reflected by some mammalian studies. Am Zool21:37–45.

Szalay FS. 2000. Evolutionary morphology. In: Delson E, TattersallI, Van Couvering JA, Brooks AS, editors. Encyclopedia ofhuman evolution and prehistory. New York: Garland Publishing.p 258–259.

Szalay FS, Bock WJ. 1991. Evolutionary theory and systematics:relationships between process and patterns. Zeit Zool Syst Evolu-Fors 29:1–39.

Tejedor MF. 2005. New fossil platyrrhines from Argentina. FoliaPrimatol 76:146–150.

Van Roosmalen MGM, van Roosmalen T. 2003. The description ofa new marmoset genus, Callibella (Callitrichinae, Primates),including its molecular phylogenetic status. Neotrop Prim 11:1–10.

Vinyard CJ, Taylor AB, Teaford MF, Glander KE, Ravosa MJ,Rossie JB, Ryan TM, Williams SH. 2011. Are we looking for loadsin all the right places? New research directions for studying themasticatory apparatus of New World monkeys. Anat Rec294:2140–2157.

Wheeler WC. 2005. Alignment, dynamic homology, and optimiza-tion. In: Albert VA, editor. Parsimony, Phylogeny and Genomics.Oxford: Oxford University Press. p 71–80.

Williams BA 2007. Comparing levels of homoplasy in the primateskeleton. J Hum Evol 52:480–489.

Wood B, Harrison T. 2011. The evolutionary context of the firsthominins. Nature 470:347–352.

Wortman JL. 1903–1904. Studies of Eocene mammalia in the MarshCollection, Peabody Museum. Part 2: Primates. Am J Sci 15:163–176; 15:399–414; 15:419–436; 16:345–368; 17:23–33; 17:133–140;17:203–214.

Youlatos D, Meldrum J. 2011. Locomotor diversification in NewWorld monkeys: running, climbing, or clawing along evolutionarybranches. Anat Rec 294:1991–2012.

1974 ROSENBERGER

evolutionary morphology, platyrrhine evolution, and...

Documents